Diffusional encounter governs the rates of many processes of biomedical importance. Experimental data on such processes now exist for enzyme/substrate, antibody/antigen, and DNA/protein systems, for example. The initial goal of this NIH project is to provide theoretical and computational methods and insights to aid in the interpretation of such data. A longer-term goal is to use theoretical and computational tools to predict the rates of diffusion-influenced processes in advance of experimental work. Such tools should ultimately be helpful in the engineering of new enzymes, chromatographic agents, etc.; indeed, experimental work published recently by Elizabeth Getzoff's group supports this conclusion. Specific aims for the next project period include the following. (1) Methods will be developed to allow for simulations of the diffusional encounter of molecules that are modeled as rigid or flexible assemblies of many spherical subunits. This will represent an advance in realism in comparison to current simulations, in which one or both molecules are modeled as rigid objects with single centers of hydrodynamic friction. (2) Methods will be developed to couple the current type of simulation of diffusional encounter (neglect of inertial dynamics; continuum electrostatics) with more detailed dynamical simulations of the encounter complexes to provide more accurate steady-state rate constants. Also, methods for calculating rates under non-steady-state conditions will be explored. (3) Improved descriptions of the electrostatic and other interactions between diffusing solutes will be developed. As these methods are developed, they will be applied to a few selected biomolecular systems for which experimental data are available, particularly Cu, Zn superoxide dismutase; triosephosphate isomerase; acetylcholinesterase; a nuclease; and monoclonal antilysozyme antibodies. Training of undergraduate, graduate, and postdoctoral students will continue to be a key aspect of this project.